1. Introduction to COVID-19 World Vaccine Adverse Reactions Dataset

2. Project work flow

3. Project methods

   3.1 Overview of important packages and verbs used

   3.2 Challenges and solutions - Load, Clean and Augment

4. Visualizations

5. Modeling

6. Conclusion and discussion

• Found on Kaggle (https://www.kaggle.com/ayushggarg/covid19-vaccine-adverse-reactions?select=2021VAERSSYMPTOMS.csv)
• Data from the Vaccine Adverse Event Reporting System (VAERS) created by the Food and Drug Administration (FDA) and Centers for Disease Control and Prevention (CDC)
• Contains 3 data sets:
  1. PATIENTS.CSV
  2. VACCINES.CSV
  3. SYMPTOMS.CSV
• Data sets connected by individual IDs (VAERS_ID)
• VAERS User Guide (https://vaers.hhs.gov/docs/VAERSDataUseGuide_November2020.pdf)
  • A guide to variable names

PATIENTS.CSV: Contains information about the individuals that received the vaccines

## # A tibble: 3 x 35
##   VAERS_ID RECVDATE  STATE AGE_YRS CAGE_YR CAGE_MO SEX   RPT_DATE   SYMPTOM_TEXT
##   <chr>    <chr>     <chr>   <dbl>   <dbl>   <dbl> <chr> <date>     <chr>       
## 1 0916600  01/01/20… TX         33      33      NA F     NA         "Right side…
## 2 0916601  01/01/20… CA         73      73      NA F     NA         "Approximat…
## 3 0916602  01/01/20… WA         23      23      NA F     NA         "About 15 m…
## # … with 26 more variables: DIED <chr>, DATEDIED <chr>, L_THREAT <chr>,
## #   ER_VISIT <chr>, HOSPITAL <chr>, HOSPDAYS <dbl>, X_STAY <chr>,
## #   DISABLE <chr>, RECOVD <chr>, VAX_DATE <chr>, ONSET_DATE <chr>,
## #   NUMDAYS <dbl>, LAB_DATA <chr>, V_ADMINBY <chr>, V_FUNDBY <chr>,
## #   OTHER_MEDS <chr>, CUR_ILL <chr>, HISTORY <chr>, PRIOR_VAX <chr>,
## #   SPLTTYPE <chr>, FORM_VERS <dbl>, TODAYS_DATE <chr>, BIRTH_DEFECT <chr>,
## #   OFC_VISIT <chr>, ER_ED_VISIT <chr>, ALLERGIES <chr>

Dimensions:

dim(patients)

## [1] 34121    35

VACCINES.CSV: Contains information about the received vaccine

## # A tibble: 3 x 8
##   VAERS_ID VAX_TYPE VAX_MANU VAX_LOT VAX_DOSE_SERIES VAX_ROUTE VAX_SITE VAX_NAME
##   <chr>    <chr>    <chr>    <chr>   <chr>           <chr>     <chr>    <chr>   
## 1 0916600  COVID19  "MODERN… 037K20A 1               IM        LA       COVID19…
## 2 0916601  COVID19  "MODERN… 025L20A 1               IM        RA       COVID19…
## 3 0916602  COVID19  "PFIZER… EL1284  1               IM        LA       COVID19…

Dimensions:

dim(vaccines)

## [1] 34630     8

SYMPTOMS.CSV: Contains information about the symptoms experienced after vaccination

## # A tibble: 3 x 11
##   VAERS_ID SYMPTOM1      SYMPTOMVERSION1 SYMPTOM2   SYMPTOMVERSION2 SYMPTOM3    
##   <chr>    <chr>                   <dbl> <chr>                <dbl> <chr>       
## 1 0916600  Dysphagia                23.1 Epiglotti…            23.1 <NA>        
## 2 0916601  Anxiety                  23.1 Dyspnoea              23.1 <NA>        
## 3 0916602  Chest discom…            23.1 Dysphagia             23.1 Pain in ext…
## # … with 5 more variables: SYMPTOMVERSION3 <dbl>, SYMPTOM4 <chr>,
## #   SYMPTOMVERSION4 <dbl>, SYMPTOM5 <chr>, SYMPTOMVERSION5 <dbl>

Dimensions:

dim(symptoms)

## [1] 48110    11

1. Load data sets (patients, vaccines, symptoms)
2. Clean each data set individually
3. Augment and merge the data sets
4. Make visualizations
5. Do modeling

Patients, vaccines and symptoms data sets:

• Multiple large files → keep them compressed as gz-files and only decompress when reading into R
• Wrong column types automatically assigned by R → manually assign appropriate column types
  
• NA strings (“NA”, “N/A”, “Unknown”, " "…) → assign NAs when loading data

CHALLENGE 1: Multiple large files

SOLUTION: Keep them compressed and only decompress when reading into R:

CHALLENGE: Wrong column types automatically assigned by R

## Warning: 241 parsing failures.
##  row          col           expected     actual         file
## 1465 BIRTH_DEFECT 1/0/T/F/TRUE/FALSE Y          <connection>
## 2742 X_STAY       1/0/T/F/TRUE/FALSE Y          <connection>
## 2807 RPT_DATE     1/0/T/F/TRUE/FALSE 2021-01-04 <connection>
## 2807 V_FUNDBY     1/0/T/F/TRUE/FALSE OTH        <connection>
## 2811 RPT_DATE     1/0/T/F/TRUE/FALSE 2021-01-04 <connection>
## .... ............ .................. .......... ............
## See problems(...) for more details.

SOLUTION: Manually assign column types

CHALLENGE: NA strings (“NA”, “N/A”, “Unknown”, " "…)

SOLUTION:

Patients data set:

• Unwanted dirty/uniformative columns → select(-c(CAGE_YR, CAGE_MO, RPT_DATE,SYMPTOM_TEXT,LAB_DATA,OFC_VISIT, ER_VISIT, X_STAY, V_FUNDBY, BIRTH_DEFECT,SPLTTYPE, RECVDATE, RECOVD, L_THREAT))

• NAs that should be interpreted as “no” → replace_na()

  • Examples: ALLERGIES and OTHER_MEDS columns

• Row duplications → distinct()

Vaccine data set:

• Contains non-COVID19 vaccines → filter(VAX_TYPE == “COVID19”)
• Row duplications → distinct()
• Duplicated IDs → add_count(VAERS_ID) %>% filter(n == 1) %>% select(-n)
• Contains vaccines of unknown manufacturer → filter(VAX_MANU != “UNKNOWN MANUFACTURER”)
• Inconsistent naming of vaccines → recode()
• Redundant and dirty columns → select(-c(VAX_NAME, VAX_LOT))

Symptoms data set:

• SYMPTOMVERSION1-5 columns are unneccessary → select(-c(SYMPTOMVERSION1, SYMPTOMVERSION2, SYMPTOMVERSION3, SYMPTOMVERSION4, SYMPTOMVERSION5))

I am aware of how horrible this table is :/

Patients data set:

• Columns containing long string descriptions → Make tidy categorical variables
  • ALLERGIES → HAS_ALLERGIES (Y/N)
  • CUR_ILL → HAS_ILLNESS (Y/N)
  • CUR_ILL → HAS_COVID (Y/N)
  • HISTORY → HAD_COVID (Y/N)
  • PRIOR_VAX → PRIOR_ADVERSE (Y/N)
  • OTHER_MEDS →
    • TAKES_ANTIINFLAMATORY (Y/N)
    • TAKES_STEROIDS (Y/N)
  • AGE_YEARS → AGE_CLASS
  • DATEDIED, VAX_DATE → DIED_AFTER
• Uninformative variable names → mutate(), case_when()
• Rows with negative values in DIED_AFTER → filter()
• Dirty, redundant and uninformative columns → select(-c(ALLERGIES, CURR_ILL, HISTORY, PRIOR_VAX, OTHER_MEDS, VAX_DATE, DATEDIED, ONSET_DATE, TODAYS_DATE))

Vaccine data set:

• No augmentation!

Symptoms data set:

• There are too many symptoms to analyze → make top_n_symptoms() function and use it to extract the top 20 occurring symptoms
• Symptoms are recorded in a way that makes later analysis difficult → turn top 20 symptoms into TRUE/FALSE columns
• Total number of symptoms needed for later analysis → mutate() to add column (N_SYMPTOMS) with total number of symptoms for each subject

Merged data sets:

• Clean and augmented patients, vaccines and symptoms data sets → inner_join() for one wide format tibble
• For certain types of analysis, we need the symptoms in a long-format → pivot_longer() to create:
  • SYMPTOM column: top 20 symptom names
  • SYMPTOM_VALUE column: TRUE/FALSE

CHALLENGE: Some columns contain long string descriptions that need to be turned into something tidy

SOLUTION: Make categorical variable

Example: ALLERGIES column:

Make categorical variable that states if patient has allergies or not:

Clean categorical HAS_ALLERGIES column:

## # A tibble: 5 x 3
##   VAERS_ID ALLERGIES                                               HAS_ALLERGIES
##   <chr>    <chr>                                                   <chr>        
## 1 0916603  Diclofenac, novacaine, lidocaine, pickles, tomatoes, m… Y            
## 2 0916604  <NA>                                                    N            
## 3 0916660  Penicillin                                              Y            
## 4 0916685  none that I am aware of                                 N            
## 5 0917437  No known allergies                                      N

Another example: OTHER_MEDS column

Detect individuals that have taken anti-inflammatory or steroid drugs before vaccine (not recommended):

Clean, categorial TAKES_ANTIINFLAMMATORY and TAKES_STEROID columns:

## # A tibble: 4 x 4
##   VAERS_ID OTHER_MEDS                           TAKES_ANTIINFLAM… TAKES_STEROIDS
##   <chr>    <chr>                                <chr>             <chr>         
## 1 0918421  1 aspirin a day 81 mg, levothyroxin… Y                 N             
## 2 0921732  Ibuprofen - PRN  States she does no… Y                 N             
## 3 0932980  Hydrocortisone 25mg daily.  Fludroc… N                 Y             
## 4 0934539  Singulair, Oxybutynin, Fosamax, Pre… N                 Y

CHALLENGE: Symptoms are recorded in a way that makes later analysis difficult

## # A tibble: 5 x 6
##   VAERS_ID SYMPTOM1           SYMPTOM2        SYMPTOM3 SYMPTOM4         SYMPTOM5
##   <chr>    <chr>              <chr>           <chr>    <chr>            <chr>   
## 1 0916618  Injection site pa… Pain            <NA>     <NA>             <NA>    
## 2 0916619  Injection site pa… Menorrhagia     <NA>     <NA>             <NA>    
## 3 0916620  Arthralgia         Chills          Headache Mobility decrea… Myalgia 
## 4 0916620  Nausea             Pain in extrem… Pyrexia  <NA>             <NA>    
## 5 0916621  Chills             Fatigue         Headache Myalgia          <NA>

SOLUTION: 20 most common symptoms are found and turned into TRUE/FALSE columns

## # A tibble: 3 x 21
##   VAERS_ID HEADACHE PYREXIA CHILLS FATIGUE PAIN  PAIN_IN_EXTREMITY NAUSEA
##   <chr>    <lgl>    <lgl>   <lgl>  <lgl>   <lgl> <lgl>             <lgl> 
## 1 0916600  FALSE    FALSE   FALSE  FALSE   FALSE FALSE             FALSE 
## 2 0916601  FALSE    FALSE   FALSE  FALSE   FALSE FALSE             FALSE 
## 3 0916602  FALSE    FALSE   FALSE  FALSE   FALSE TRUE              FALSE 
## # … with 13 more variables: DIZZINESS <lgl>, MYALGIA <lgl>,
## #   INJECTION_SITE_ERYTHEMA <lgl>, INJECTION_SITE_PRURITUS <lgl>,
## #   INJECTION_SITE_SWELLING <lgl>, INJECTION_SITE_PAIN <lgl>, ARTHRALGIA <lgl>,
## #   DYSPNOEA <lgl>, VOMITING <lgl>, PRURITUS <lgl>, DEATH <lgl>, RASH <lgl>,
## #   ASTHENIA <lgl>

## # A tibble: 7 x 6
##   term           estimate std.error statistic  p.value odds_ratio
##   <chr>             <dbl>     <dbl>     <dbl>    <dbl>      <dbl>
## 1 (Intercept)    -9.34      0.161    -58.0    0         0.0000876
## 2 SEXM            0.924     0.0573    16.1    2.18e-58  2.52     
## 3 AGE_YRS         0.0915    0.00207   44.2    0         1.10     
## 4 HAS_ALLERGIESY -0.100     0.0608    -1.65   9.82e- 2  0.904    
## 5 HAS_ILLNESSY    1.10      0.0664    16.6    6.60e-62  3.01     
## 6 HAS_COVIDY     -0.117     0.148     -0.791  4.29e- 1  0.890    
## 7 HAD_COVIDY      0.00915   0.193      0.0474 9.62e- 1  1.01

## # A tibble: 20 x 6
##   term          estimate std.error statistic  p.value odds_ratio
##   <chr>            <dbl>     <dbl>     <dbl>    <dbl>      <dbl>
## 1 (Intercept)     -2.01     0.0287    -70.1  0             0.134
## 2 HEADACHETRUE    -1.67     0.156     -10.7  7.92e-27      0.188
## 3 PYREXIATRUE     -0.429    0.112      -3.82 1.34e- 4      0.651
## 4 CHILLSTRUE      -1.21     0.171      -7.11 1.17e-12      0.298
## 5 FATIGUETRUE     -0.367    0.115      -3.19 1.41e- 3      0.693
## 6 PAINTRUE        -0.913    0.153      -5.98 2.17e- 9      0.401
## 7 NAUSEATRUE      -0.621    0.139      -4.46 8.17e- 6      0.538
## 8 DIZZINESSTRUE   -2.17     0.193     -11.2  2.87e-29      0.114
## # … with 12 more rows

## # A tibble: 20 x 9
##   SYMPTOM  estimate std.error statistic p.value conf.low conf.high odds_ratio
##   <chr>       <dbl>     <dbl>     <dbl>   <dbl>    <dbl>     <dbl>      <dbl>
## 1 HEADACHE  -0.170     0.0954    -1.79   0.0742   -0.361    0.0133      0.843
## 2 PYREXIA    0.0734    0.0967     0.760  0.448    -0.120    0.259       1.08 
## 3 CHILLS    -0.0727    0.103     -0.703  0.482    -0.280    0.126       0.930
## 4 FATIGUE    0.0226    0.102      0.221  0.825    -0.183    0.219       1.02 
## 5 PAIN       0.0190    0.106      0.179  0.858    -0.194    0.222       1.02 
## # … with 15 more rows, and 1 more variable: identified_as <chr>

Chi-squared contingency table tests

Important verbs and tools used:

• prcomp()
• kmeans()
• tidymodels: (used for what?)